Effects of 2-m air temperature assimilation and a new near-surface observation operator on the NCEP gridpoint statistical-interpolation system

In this study, we assimilate 2-m air temperature data with the National Centers for Environmental Prediction (NCEP) regional Gridpoint Statistical Interpolation (GSI) using the WRF-NMM model forecast as a first guess. Single time analysis experiments are conducted to test the impact of 2-m air temperature data on the analysis system and the results are compared with the control run without using 2-m air temperature data. The effort is focused on understanding the characteristics of observation innovations of the 2-m air temperature data. Modifications to background errors and a simple test of nonlinear quality control are also considered. The incorporation of a comprehensive near-surface observation operator based on Monin-Obukhov similarity theory is described and tested for possible operational use with the NCEP regional GSI system. The results from this new forward operator are compared with those from the existing simple forward operator. According to the results, mesonet 2-m temperature data were found to have a considerable amount of outliers compared with other 2- m temperature data. The nighttime western and central US domains indicated a model warm bias. Stations with large innovations are distributed uniformly in the nighttime western and central domains, while they are mainly located in the large cities in the daytime eastern domain. The statistical analysis of observation innovations showed that introduction of the new forward model can reduce root-mean-square errors in observation increment statistics. The results of a short assimilation experiment indicate that the new forward operator can be employed as a short-term strategy for near-surface data assimilation in the NCEP.